schizophrenia We invite articles on the use of computational algorithms to predict the onset or course of psychosis. Bridging advanced analytics and clinical practice, this collection focuses on reproducible and interpretable models ranging from neuroimaging to language processing to enhance early detection and intervention of mental disorders.
This collection focuses on innovative machine learning applications for predicting the onset or progression of psychotic disorders. We welcome research using supervised, unsupervised, or deep learning techniques applied to clinical, neuroimaging, electrophysiological, genetic, linguistic, or multimodal datasets to improve risk stratification and early diagnosis. The focus is on integrating transparent and reproducible algorithms, external validation, and natural language processing of clinical records and audio samples. Research bridging computational methods and clinical utility that demonstrates model interpretability, cost-effectiveness, or deployment into early intervention services is strongly encouraged.
Contributions should detail methodological advances and practical implementations of machine learning models for early psychosis. Suitable topics include feature engineering from neurobiological and psychosocial data, validating predictive pipelines across cohorts, NLP for patient narratives, integrating digital phenotyping (e.g., smartphone and social media data), and translational frameworks for clinical implementation.
Topics of interest include, but are not limited to:
- Deep learning of MRI and EEG for psychosis risk
- NLP analysis of clinical interviews
- Selection of genetic and epigenetic features
- Digital phenotyping using mobile data
- Cross-site validation of predictive models
This collection supports and expands research related to: SDG 3 (Health and Wellbeing).
